System and method for adaptive cell search based on artificial intelligence model of an electronic device

Abstract

A method for selecting a target cell by a user equipment (UE), includes: detecting a movement of the UE based on a change in a location of the UE and a change in a measured signal power of a serving cell; determining personalized data of a user based on a plurality of user contextual parameters; predicting a destination of the user, and one or more target cells across a path of the user to the destination based on the determined personalized data of the user and the detected movement of the UE; and selecting the target cell from the one or more target cells, based on an artificial intelligence (AI) model. The AI model is configured to update weights of one or more network quality parameters.

What is claimed is: 
     
         1 . A method for selecting a target cell by a user equipment (UE), the method comprising:
 detecting a movement of the UE based on a change in a location of the UE and a change in measured signal power of a serving cell;   determining personalized data of a user based on a plurality of user contextual parameters;   predicting a destination of the user, and   one or more target cells across a path of the user to the destination based on the determined personalized data of the user and the detected movement of the UE; and   selecting the target cell from the one or more target cells, based on an artificial intelligence (AI) model,   wherein the AI model is configured to update a weight for each of one or more network quality parameters, based on at least one of:
 the determined personalized data, 
 information on the one or more target cells, and 
 a mapping of the target cell from the one or more target cells and the serving cell across the path of the user to the destination. 
   
     
     
         2 . The method of  claim 1 , further comprising:
 tracking the change in the location of the UE and a geographical location of the UE, based on sensor data;   obtaining information about the measured signal power of the serving cell from a network;   determining the change in the measured signal power of the serving cell by comparing the measured signal power of the serving cell with a threshold value, the measured signal power of the serving cell being lower than the threshold value; and   detecting the movement of the UE based on the determined change in the measured signal power of the serving cell and the tracked geographical location of the UE,   wherein the tracked geographical location of the UE is stored in a memory.   
     
     
         3 . The method of  claim 1 , further comprising:
 obtaining the plurality of user contextual parameters from at least one of the UE and the network; and   determining the personalized data of the user based on the obtained plurality of user contextual parameters and the movement of the UE,   wherein the plurality of user contextual parameters comprises at least one of:
 call schedules of the user, 
 calling preferences of the user, 
 an ongoing call type of the user, 
 an estimated battery life, 
 a throughput value of the UE, 
 a preference of the path of the user, 
 historical activity data of the user, and 
 a last location of the user. 
   
     
     
         4 . The method of  claim 1 , further comprising fetching information about one or more target cells among the one or more target cells, and the one or more network quality parameters of the one or more target cells based on the prediction,
 wherein the predicted destination of the user comprises coordinates associated with the destination of the user, and the one or more target cells.   
     
     
         5 . The method of  claim 4 , wherein the information about one or more target cells comprises at least one of a frequency, Reference Signal Received Power (RSRP) of the one or more target cells, and Physical Cell identification Information (PCI ID). 
     
     
         6 . The method of  claim 4 , wherein the one or more network quality parameters comprises at least one of a network load, radio frequency (RF) link imbalance, radio access technology (RAT) information, and a size of the target cell. 
     
     
         7 . The method of  claim 4 , further comprising:
 identifying one or more triggering points at instances where the movement of the UE is detected, the one or more triggering points corresponding to the tracked geographical location;   obtaining event context data for each of the one or more target cells and serving cells at the identified one or more triggering points, the event context data comprising the information about one or more target cells;   obtaining a location of each of the identified one or more triggering points based on sensor data;   calculate the one or more network quality parameters of the one or more target cells at the identified one or more triggering points;   performing the mapping of the target cell and the serving cell at a triggering point of the identified one or more triggering points to generate a log of event context data and the location of each of the identified one or more triggering points; and   storing the log of event context data along with the location at a memory of the UE.   
     
     
         8 . The method of  claim 1 , further comprising:
 inputting, in the AI model, a hot encoding of the one or more network quality parameters as training data for training the AI model;   generating the hot encoding of the one or more network quality parameters based on an encoding of the one or more network quality parameters along with the path of the user to the destination based on the movement of the UE; and   training the AI model and obtaining a trained AI model by assigning the weight to each of the one or more network quality parameters based on the inputted hot encoding of the one or more network quality parameters.   
     
     
         9 . The method of  claim 7 , further comprising training of the AI model based on at least one of the trained AI model, information on the target cell, the one or more network quality parameters, and personalized data,
 wherein the training of the AI model comprises:
 calculating a threshold value for each of the personalized data of the user by using a regression model; 
 calculating a loss value based on the generated log of event context data and the calculated threshold value; 
 applying the calculated loss value on the trained AI model; and 
 updating the weight for each of the one or more network quality parameters based on the calculated threshold value. 
   
     
     
         10 . A user equipment (UE) for selecting a target cell, comprising:
 one or more processors configured to:
 detect a movement of the UE based on a change in a location of the UE and a change in measured signal power of a serving cell; 
 determine personalized data of a user based on a plurality of user contextual parameters; 
 predict a destination of the user, and one or more target cells across a path of the user to the destination based on the personalized data and the detected movement of the UE; and 
 select the target cell from one or more target cells based on a trained AI model that updates a weight for each of the one or more network quality parameters based on at least one of the personalized data, 
   information about one or more target cells, and   a mapping of the target cell from the one or more target cells and the serving cell,   across the path of the user to the destination based on training data and the prediction.   
     
     
         11 . The UE of  claim 10 , wherein the one or more processors are further configured to:
 track for the change in the location, a geographical location of the UE, based on sensor data;   obtain information about the measured signal power of the serving cell from a network;   determine the change in the measured signal power of the serving cell by comparing the measured signal power of the serving cell with a threshold value, the measured signal power of the serving cell being lower than the threshold value; and   detect the movement of the UE based on the determination of the change in the measured signal power of the serving cell and the tracked geographical location of the UE,   wherein the tracked geographical location of the UE is stored in a memory.   
     
     
         12 . The UE of  claim 11 , wherein the one or more processors are further configured to:
 obtain the plurality of user contextual parameters from at least one of the UE and the network, and   determine the personalized data of the user based on the obtained plurality of user contextual parameters and the detected movement of the UE,   wherein the plurality of user contextual parameters comprises at least one of:
 call schedules of the user, 
 calling preferences of the user, 
 ongoing call type of the user, 
 estimated battery life, 
 a throughput value of the UE, 
 a preference of the path of the user, 
 historical activity data of the user, and 
 a last location of the user. 
   
     
     
         13 . The UE of  claim 10 , wherein the one or more processors are further configured to:
 fetch at least one of (i) information about one or more target cells among the one or more target cells and (ii) the one or more network quality parameters of the one or more target cells, based on the prediction,   wherein the predicted destination of the user comprises coordinates associated with the destination of the user and the one or more target cells.   
     
     
         14 . The UE of  claim 10 , wherein the information about one or more target cells comprises at least one of a frequency, Reference Signal Received Power (RSRP) of the one or more target cells, and Physical Cell identification Information (PCI ID). 
     
     
         15 . The UE of  claim 10 , wherein the one or more network quality parameters comprises at least one of a network load, radio frequency (RF) link imbalance, radio access technology (RAT) information, and a size of the target cell. 
     
     
         16 . The UE of  claim 13 , wherein the one or more processors are further configured to:
 identifying one or more triggering points at instances where the movement of the UE is detected, the one or more triggering points corresponding to the tracked geographical location;   obtaining event context data respective of the one or more target cells and one or more serving cells at the identified one or more triggering points, the event context data comprising the information about one or more target cells;   obtaining a location of each of the identified one or more triggering points based on sensor data;   calculate the one or more network quality parameters of the one or more target cells at the identified one or more triggering points;   performing the mapping of the target cell and a serving cell at a triggering point of the identified one or more triggering points to generate a log of event context data and the location of each of the identified one or more triggering points; and   storing the log of event context data along with the location at a memory of the UE.   
     
     
         17 . The UE of  claim 16 , wherein the one or more processors are further configured to:
 input, in the AI model, a hot encoding of the one or more network quality parameters as the training data for training the AI model;   generate the hot encoding of the one or more network quality parameters based on an encoding of the one or more network quality parameters along with the path of the user to the destination based on the movement of the UE; and   train the AI model and obtain a trained AI model by assigning the weight to each of the one or more network quality parameters based on the inputted hot encoding of the one or more network quality parameters.   
     
     
         18 . The UE of  claim 17 , wherein the one or more processors are further configured to perform a training of the AI model based on the trained AI model, target cell information, the one or more network quality parameters and personalized data,
 wherein the training of the AI model comprises:
 calculate a threshold value for each of the personalized data of the user by using a regression model; 
 calculate a loss value based on the generated log and the calculated threshold value; 
 apply the calculated loss value on the trained AI model; and 
 update the weight for each of the one or more network quality parameters based on the calculated threshold value.